>The exact same flaw exists in the design of PGP/GPG and whenever it comes up in that context it's a reason to throw GPG into the garbage disposal.
I literally never heard of this. There are problems with PGP (eg. no forward secrecy, non-reputability, unencrypted headers) but "your counterparty could be compromised" isn't one of them.
I think the reference is to the idea that a correspondent might do a unencrypted CC of a message that contains previously encrypted text as per this infamous anti-PGP rant:
I'm referring to "your counterparty can hit reply-all and forget to encrypt" which is a mistake in the same category as "your counterparty might have backups enabled", i.e. it's easy to misuse in a way that ends up defeating secrecy.
I literally never heard of this. There are problems with PGP (eg. no forward secrecy, non-reputability, unencrypted headers) but "your counterparty could be compromised" isn't one of them.